Rit (Vísindafélag Íslendinga) - 01.06.1984, Side 266

262
PÁU. IMSI.AND
subtracting from it the wehrlite minerals, in order to produce a Mg poorer
ankaramite is a rather crude operation, not least because the wehrlite
minerals are partly out oí' equilibrium with the ankaramite composition.
Nevertheless, this illustrates the effectiveness of the wehrlite fractionation
and shows clearly the dominating iníluence of the clinopyroxene over olivine
in the early evolution of this rock suite.
Except for the wehrlite minerals, the ankaramites contain phenocrysts of
less magnesian olivine and zoned titan-salites, which are the products of the
crystallization in the ankaramitic liquids after the wehrlite crystallization.
This ankaramitic crystallization is assumed to happen after the liquid leaves
the wehrlite crystallization conditions and starts to move towards the
surface but before it enters the plagioclase stability field.
In Tables 40 and 41 are shown the results of crystal fractionation within
the ankaramitic rocks ofjan Mayen according to least-squares calculations.
In the case of the wehrlite crystallization the “ankaramitic” liquids are
fractionated by removal of a chromian diopside and a Fo rich olivine. The
minerals chosen are minerals contained by the ankaramites. The mineral
compositions represent an average composition for the relevant chromian
diopsides and highly magnesian olivines of the wehrlite crystallization
category. They are thus in equilibrium with an undefined ankaramitic
liquid. By following the liquid changes, by changing the mineral composi-
tions step by step, and thus more closely tracing the equilibrium conditions,
a better fit could be obtained. In spite of the rather poor equilibrium
approximation used, the results obtained are satisfactory. Table 40 shows
the compositions of the fractionating minerals, the parental liquid — the
most magnesian ankaramite (Jan 166) — and four pairs of analysed and
calculated derivatives of successively decreasing Mg content and primitive-
ness. In these cases 36 to 58 per cent of the minerals must be subtracted and
of this fractionate 75 to 80 per cent are chromian diopside. The total fit
between the observed and calculated derivatives is tolerable. Undertaking
this operation in a steplike manner — subtracting the minerals from the
most primitive rock composition to produce a slightly less primitive one and
then the third one from the second one, and so on — as shown in Table 41,
gives a better fit. The result has the same characteristics. Clinopyroxene still
dominates the fractionate and amounts to 64 to 82 per cent. The total
amount of subtracted minerals is in this case greater. It is around 20 per
cent in each step and amounts to 78 per cent in fractionating from Jan 166,
through Jan 12, 13 and 10 to Jan 65, while fractionating between these two
compositions in one step needs only 58 per cent subtracted minerals.
In both these cases the elements showing the poorest fit are Al, Fe, and
Cr. As Cr enters a chrome-spinel which may contain up to more than 50 per
cent Cr203 by weight, as well as the chromian diopside in these rocks, a very
small amount of this mineral has a major efl'ect. Such minor amounts of a
trace element mineral are almost impossible to account for correctly in